ISUP Call Setup With ACM and ANM Explained
In an ISDN User Part (ISUP) call, signalling messages move through the SS7 network before speech or media is established. The Initial Address Message (IAM) starts the attempt, while the Address Complete Message (ACM) and Answer Message (ANM) mark two important stages in the call’s progress.
These messages are easy to confuse because both ACM and ANM travel in the backward direction. ACM usually means that routing is complete and the destination is being alerted. ANM means the called party has answered, so the call has entered the conversation state.
For engineers working with Australian fixed, mobile, and international interconnects, understanding these states helps when tracing delayed call setup, incorrect charging, early media, or failed number translation. A clear message sequence also makes packet captures and SS7 test results much easier to interpret.
The ISUP Call Setup Sequence
The originating exchange sends an IAM after collecting enough digits to route the call. The IAM carries parameters such as the called party number, calling party information, nature of connection indicators, and forward call indicators. Intermediate signalling points may relay or process this message before it reaches the destination exchange.
When the terminating side has analysed the number and can continue call handling, it returns an ACM. This response confirms that the address information is complete from the receiving network’s perspective. It does not confirm that a person has answered.
The destination exchange then alerts the subscriber, which may produce ringing at a handset or an equivalent alerting state in a service platform. During this period, the originating side may provide ringback tone to the caller, depending on network design and whether tones are generated locally or passed through the interworking path.
What ACM Actually Tells The Network
ACM is best understood as an alerting or call-progress indication. It tells the originating exchange that the call has reached a valid stage beyond basic digit collection. The terminating network may still be waiting for the subscriber to answer, for a network service to respond, or for another application decision.
A network can return ACM while the called number is ringing, while a mobile service is locating a handset, or while an interworking gateway is completing a controlled setup. The exact user experience depends on the exchange, service logic, and bearer arrangement.
In Australia, this distinction matters across long-distance links between Sydney, Melbourne, Brisbane, and Perth, where multiple carriers or transit providers may be involved. A caller hearing ringback does not prove that ANM has been sent; the tone may be generated by an originating switch after receiving ACM.
ANM And The Answered State
ANM is sent in the backward direction when the terminating party answers. It changes the call from an alerting condition to an answered condition. The originating exchange can then begin answer-related accounting and treat the connection as active for ordinary voice service.
A useful simplified sequence is IAM, ACM, ANM, followed later by Release (REL) and Release Complete (RLC). REL indicates that one side wants to clear the call, while RLC confirms that the release procedure has completed. If ANM never appears, the call should not be treated as answered merely because ACM was received.
Charging behaviour is governed by carrier and service policy, so ANM should not be treated as a universal billing rule in every modern platform. Still, it remains a key diagnostic marker when checking answer supervision, post-dial delay, and unexpected charge events.
Timers, Failures, And Interworking
ISUP timers protect the call state machine from waiting indefinitely. If a response such as ACM or ANM does not arrive within the expected interval, the originating exchange may clear the call or invoke a recovery procedure. Timer names and values vary by implementation and standards profile, so traces should always be compared with the relevant equipment documentation.
Common faults include an absent ACM, an ANM sent too early, duplicated messages, or a release caused by incompatible signalling parameters. Number portability databases, international prefixes, and mobile routing can add further complexity. An Australian call to a number hosted in another carrier’s network may pass through several translation and interconnect stages before reaching the subscriber.
Engineers using SS7 training materials can connect these message states with MTP3 routing, signalling point codes, circuit identification codes, and SIGTRAN transport over IP. That wider context is important because a correct ISUP message can still fail to reach its destination if the underlying signalling route is unavailable.
Reading A Trace Efficiently
A trace should be read in time order and in relation to the circuit or logical bearer. Start with the IAM, verify the destination address and circuit identification, then find the first backward response. ACM should be checked for its presence and timing, while ANM should be correlated with the actual answer event.
Useful checks include:
- Confirm the IAM contains the expected called number and routing indicators.
- Match ACM and ANM to the correct circuit identification code.
- Measure the interval from IAM to ACM and from ACM to ANM.
- Check whether REL appears before ANM or after the answered state.
The message direction is equally important. IAM travels from the originating exchange towards the destination, while ACM, ANM, REL, and RLC may travel in the reverse direction depending on which side initiates clearing. A simple signalling ladder often reveals a fault faster than a large protocol capture.
When documenting test calls, include the access network, transit carrier, destination type, and local time. Conditions can differ between a business service in Melbourne, a mobile subscriber in regional Queensland, and an international route from Perth. Even a travel-oriented geographic perspective can reinforce why distance, routing boundaries, and local network architecture matter when interpreting real-world call paths.
ACM And ANM At A Glance
| Message | Direction | Main meaning | Typical user state | What it does not prove |
|---|---|---|---|---|
| IAM | Forward | Starts call setup | Digits have been collected | That the destination is reachable |
| ACM | Backward | Address is complete and progress can continue | Alerting or call progressing | That the called party answered |
| ANM | Backward | Called party or service answered | Conversation state | That media quality is good |
| REL | Either direction | Requests call clearing | Call ending | That release is fully complete |
| RLC | Either direction | Confirms release completion | Resources being returned | That every application session has closed |
The central distinction is simple: ACM represents progress towards connection, while ANM represents answer supervision. Analysts should confirm both the message sequence and the timing before diagnosing a call as connected.
Remembering The Two Answer States
ACM belongs to the alerting phase. It confirms that the destination side has accepted the address and can continue handling the call, but ringing or service processing may still be underway.
ANM belongs to the answered phase. It signals that the call has crossed into active conversation handling, making it the decisive marker when investigating answer timing, charging triggers, or an unexpectedly short call.
The detail to remember is that ACM says “the call is progressing,” whereas ANM says “the call has been answered.”
